BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20210402T160556Z
LOCATION:Track 10
DTSTART;TZID=America/New_York:20201112T174000
DTEND;TZID=America/New_York:20201112T180500
UID:submissions.supercomputing.org_SC20_sess218_ws_pawatm109@linklings.com
SUMMARY:HOOVER: Leveraging OpenSHMEM for High Performance, Flexible, Strea
 ming Graph Applications
DESCRIPTION:Workshop\n\nHOOVER: Leveraging OpenSHMEM for High Performance,
  Flexible, Streaming Graph Applications\n\nGrossman, Pritchard, Poole, Sar
 kar\n\nAs the adoption of streaming graph applications increases in the fi
 nance, defense, social and other industries, so does the size, velocity, i
 rregularity and complexity of these streaming graphs. Most existing framew
 orks for processing streaming, dynamic graphs are too limited in scalabili
 ty or functionality to efficiently handle these increasingly massive graph
 s. Many frameworks are either built for shared memory platforms only (limi
 ting the size of the graph that can be stored) or for distributed platform
 s, but run on slow, high overhead, interpreted and bulk synchronous platfo
 rms.\n\nThis paper introduces HOOVER, a high performance streaming graph m
 odeling and analysis framework built from scratch to scale on high perform
 ance systems and extremely dynamic graphs. HOOVER offers similar APIs to p
 revious streaming graph frameworks, but sits on top of a high performance 
 runtime system designed for modern supercomputers. HOOVER leverages an eve
 ntually-consistent consistency model to improve scalability, and offers a 
 number of unique features to users. On micro-benchmarks, HOOVER is shown t
 o be comparable or faster than existing high performance and distributed g
 raph frameworks. Using mini-apps, we also show that HOOVER easily scales t
 o 2,048 PEs on more realistic applications.\n\nRegistration Category: Work
 shop Reg Pass
END:VEVENT
END:VCALENDAR